ELVM Compiler Infrastructure
ELVM is similar to LLVM but dedicated to Esoteric Languages. This project consists of two components - frontend and backend. Currently, the only frontend we have is a modified version of 8cc. The modified 8cc translates C code to an internal representation format called ELVM IR (EIR). Unlike LLVM bitcode, EIR is designed to be extremely simple, so there's more chance we can write a translator from EIR to an esoteric language.
Currently, there are 56 backends:
- Awk (by @dubek)
- Bash
- Befunge
- Binary Lambda Calculus (by @woodrush)
- Brainfuck
- C
- C++14 constexpr (compile-time) (by @kw-udon)
- C++ Template Metaprogramming (compile-time) (by @kw-udon) (WIP)
- C# (by @masaedw)
- C-INTERCAL
- CMake (by @ooxi)
- CommonLisp (by @youz)
- Conway's Game of Life (via QFTASM) (by @woodrush)
- Crystal (compile-time) (by @MakeNowJust)
- Emacs Lisp
- F# (by @masaedw)
- Forth (by @dubek)
- Fortran (by @samcoppini)
- Go (by @shogo82148)
- Grass (by @woodrush)
- HeLL (by @esoteric-programmer)
- J (by @dubek)
- Java
- JavaScript
- Kinx (by @Kray-G)
- Lambda calculus (by @woodrush)
- Lazy K (by @woodrush)
- LLVM IR (by @retrage)
- LOLCODE (by @gamerk)
- Lua (by @retrage)
- Octave (by @inaniwa3)
- Perl5 (by @mackee)
- PHP (by @zonuexe)
- Piet
- Python
- Ruby
- Scheme syntax-rules (by @zeptometer)
- Scratch3.0 (by @algon-320)
- SQLite3 (by @youz)
- SUBLEQ (by @gamerk)
- Swift (by @kwakasa)
- Tcl (by @dubek)
- TeX (by @hak7a3)
- TensorFlow (WIP)
- Turing machine (by @ND-CSE-30151)
- Unlambda (by @irori)
- Universal Lambda (by @woodrush)
- Vim script (by @rhysd)
- WebAssembly (by @dubek)
- WebAssembly System Interface (by @sanemat)
- Whirl by (@samcoppini)
- W-Machine by (@jcande)
- Whitespace
- arm-linux (by @irori)
- i386-linux
- sed
The above list contains languages which are known to be difficult to program in, but with ELVM, you can create programs in such languages. You can easily create Brainfuck programs by writing C code for example. One of interesting testcases ELVM has is a tiny Lisp interpreter. The all above language backends are passing the test, which means you can run Lisp on the above languages.
Moreover, 8cc and ELVM themselves are written in C. So we can run a C compiler written in the above languages to compile the ELVM's compiler toolchain itself, though such compilation takes long time in some esoteric languages.
A demo site
http://shinh.skr.jp/elvm/8cc.js.html
As written, ELVM toolchain itself runs on all supported language backends. The above demo runs ELVM toolchain on JavaScript (thus slow).
Example big programs
- 8cc in JavaScript
- 8cc in Brainfuck
- 8cc in Unlambda
- Lisp in Piet
- Lisp in C-INTERCAL
- 8cc in Befunge
- 8cc in Whitespace
ELVM internals
ELVM IR
- Harvard architecture, not Neumann (allowing self-modifying code is hard)
- 6 registers: A, B, C, D, SP, and BP
- Ops: mov, add, sub, load, store, setcc, jcc, putc, getc, and exit
- Psuedo ops: .text, .data, .long, and .string
- mul/div/mod are implemented by _builtin*
- No bit operations
- No floating point arithmetic
- sizeof(char) == sizeof(int) == sizeof(void*) == 1
- The word-size is backend dependent, but most backend uses 24bit words
- A single programming counter may contain multiple operations
See ELVM.md for more detail.
Directories
shinh/8cc's eir branch is the frontend C compiler.
ir/ directory has a parser and an interpreter of ELVM IR. ELVM IR has
target/ directory has backend implementations. Code in this directory uses the IR parser to generate backend code.
libc/ directory has an incomplete libc implementation which is necessary to run tests.
Notes on language backends
Brainfuck
Running a Lisp interpreter on Brainfuck was the first motivation of this project (bflisp). ELVM IR is designed for Brainfuck but it turned out such a simple IR could be suitable for other esoteric languages.
As Brainfuck is slow, this project contains a Brainfuck interpreter/compiler in tools/bfopt.cc. You can also use other optimized Brainfuck implementations such as tritium. Note you need implementations with 8bit cells. For tritium, you need to specify `-b' flag.
Unlambda
This backend was contributed by @irori. See also 8cc.unl.
This backend is tested with @irori's interpreter. tools/rununl.sh automatically downloads it.
C-INTERCAL
This backend uses 16bit registers and address space, though ELVM's standard is 24bit. Due to the lack of address space, you cannot compile large C programs using 8cc on C-INTERCAL.
This backend won't be tested by default because C-INTERCAL is slow. Use
$ CINT=1 make i
to run them. Note you may need to adjust tools/runi.sh.
You can make faster executables by doing something like
$ cp out/fizzbuzz.c.eir.i fizzbuzz.i && ick fizzbuzz.i
$ ./fizzbuzz
But compilation takes much more time as it uses gcc instead of tcc.
Piet
This backend also has 16bit address space. There's the same limitation as C-INTERCAL's.
This backend won't be tested by default because npiet is slow. Use
$ PIET=1 make piet
to run them.
Befunge
BefLisp, which translates LLVM bitcode to Befunge, has very similar code. The interpreter, tools/befunge.cc is mostly Befunge-93, but its address space is extended to make Befunge-93 Turing-complete.
Whitespace
This backend is tested with @koturn's Whitespace implementation.
Emacs Lisp
This backend is somewhat more interesting than other non-esoteric backends. You can run a C compiler on Emacs:
- M-x load-file tools/elvm.el
- open test/putchar.c (or write C code without #include)
- M-x 8cc
- Now you'll see ELVM IR. You need to prepend a backend name (`el' for example) as the first line.
- M-x elc
- M-x eval-buffer
- M-x elvm-main
Vim script
This backend was contributed by @rhysd. You can run a C compiler on Vim:
- Open test/hello.c (or write your C code)
:source /path/to/out/8cc.vim
- Now you can see ELVM IR in the buffer
- Please prepend a backend name (
vim
for Vim) to the first line :source /path/to/out/elc.vim
- You can see Vim script code as the compilation result in current buffer
- You can
:source
to run the code
You can find more descriptions and released vim script in 8cc.vim.
TeX
This backend was contributed by @hak7a3. See also 8cc.tex.
C++14 constexpr (compile-time)
This backend was contributed by @kw-udon. You can find more descriptions in constexpr-8cc.
sed
This backend is very slow so only limited tests run by default. You can run them by
$ FULL=1 make sed
but it could take years to run all tests. I believe C compiler in sed works, but I haven't confirmed it's working yet. You can try Lisp interpreter instead:
$ FULL=1 make out/lisp.c.eir.sed.out.diff
$ echo '(+ 4 3)' | time sed -n -f out/lisp.c.eir.sed
This backend should support both GNU sed and BSD sed, so this backend is more portable than sedlisp, though much slower. Also note, due to limitation of BSD sed, programs cannot output non-ASCII characters and NUL.
HeLL
This backend was contributed by @esoteric-programmer. HeLL is an assembly language for Malbolge and Malbolge Unshackled. Use LMFAO to build the Malbolge Unshackled program from HeLL. This backend won't be tested by default because Malbolge Unshackled is extremely slow. Use
$ HELL=1 make hell
to run them. Note you may need to adjust tools/runhell.sh.
This backend does not support all 8-bit characters on I/O, because I/O of Malbolge Unshackled uses Unicode codepoints instead of single bytes in getc/putc calls. Further, the Malbolge Unshackled interpreter automatically converts newlines read from stdin, which cannot be revert in a platform independent way. The backend reverts/converts newlines from input to Linux encoding and applies modulo 256 operations to all input and output, but it cannot compensate the issues this way. You should limit I/O to ASCII characters in order to avoid unexpected behaviour or crashes.
This backend may be replaced by a Malbolge Unshackled backend in the future.
TensorFlow
Thanks to control flow operations such as tf.while_loop and tf.cond, a TensorFlow's graph is Turing complete. This backend translates EIR to a Python code which constructs a graph which is equivalent to the source EIR. This backend is very slow and uses a huge amount of memory. I've never seen 8cc.c.eir.tf works, but lisp.c.eir.tf does work. You can test this backend by
$ TF=1 make tf
TODO: Reduce the size of the graph and run 8cc
Scratch 3.0
Scratch is a visual programming language.
Internally, a Scratch program consists of a JSON that represent the program and some resources such as images or sounds. They are zip-archived and you can import/export them from project page (Create new one from here).
You can use tools/gen_scratch_sb3.sh
to generate complete project files from output of this backend,
and tools/run_scratch.js
to execute programs from command line (npm 'scratch-vm' package is required).
You can try "fizzbuzz_fast" sample from here.
test/basic.eir
)
Example (for First, generate scratch project.
$ ./out/elc -scratch3 test/basic.eir > basic.scratch3
$ ./tools/gen_scratch_sb3.sh basic.scratch3
$ ls basic.scratch3.sb3
basic.scratch3.sb3
Execute it from Web browser
- Visit https://scratch.mit.edu/projects/editor.
- Click a menu item: "File".
- Click "Load from your computer".
- Select and upload the generated project file:
basic.scratch3.sb3
. - Wait until the project is loaded. (It takes a long time for a hevy project.)
- Click the "Green Flag"
From the Web editor, to input special characters (LF, EOF, etc.) you have to input them explicitly by following:
special character | representation |
---|---|
LF | \n |
EOF | \0 |
other character with codepoint XXX (decimal) | \dXXX |
Note that: the escape character is \
(U+FF3C) not \
.
For normal ASCII characters, you can just put them into the input field.
Execute it from command line
- First install the npm package "scratch-vm" under the
tools
directory :
$ cd tools
$ npm install scratch-vm
- Run it with
tools/run_scratch.js
:
$ echo -n '' | nodejs ./run_scratch.js ../basic.scratch3.sb3
!!@X
Conway's Game of Life
This backend was contributed by @woodrush based on QFTASM. See tools/qftasm/README.md for its details. Further implementation details are described in the Lisp in Life project.
Binary Lambda Calculus
This backend was contributed by @woodrush. Implementation details are described in the LambdaVM and lambda-8cc repositories.
The output of this backend is an untyped lambda calculus term written in binary lambda calculus notation. The output program runs on the IOCCC 2012 "Most Functional" interpreter written by @tromp. The program runs on the byte-oriented mode which is the default mode.
This backend outputs a sequence of 0/1s written in ASCII. This bit stream must be packed into a byte stream before passing it to the interpreter, which can be done using tools/packbits.c. Please see tools/runblc.sh for usage details.
This backend is tested with the interpreter uni, a fast implementation of the "Most Functional" interpreter written in C++ by @melvinzhang. This interpreter significantly speeds up the running time of large programs such as 8cc.c. tools/runblc.sh automatically clones and builds uni via tools/runblc.sh when the tests are run.
Lambda Calculus
This backend was contributed by @woodrush.
This backend outputs an untyped lambda calculus term written in plain text, such as \x.(x x)
.
The I/O model used in this backend is identical to the one used in the Binary Lambda Calculus backend. The backend's output program is a lambda calculus term that takes a string as an input and returns a string. Here, strings are encoded into lambda calculus terms using Scott encoding and Church encoding, so the entire computation only consists of the beta-reduction of lambda calculus terms. Further implementation details are described in the LambdaVM and lambda-8cc repositories. Note that the backend's output program is assumed to be evaluated using a lazy evaluation strategy.
This backend is tested with the interpreter uni,
written by @melvinzhang.
The blc tool written by @tromp is also used to convert plain text lambdas into binary lambda calculus notation, the format accepted by uni
.
Both tools are automatically cloned and built via tools/runlam.sh when the tests are run.
Lazy K
The Lazy K backend was contributed by @woodrush. Implementation details are described in the LambdaVM and lambda-8cc repositories.
This backend is tested with the Lazy K interpreter lazyk written by @irori.
Interactive programs require the -u
option which disables standard output buffering, used as lazyk -u [input file]
.
The interpreter is automatically cloned and built via tools/runlazy.sh when the tests are run.
Universal Lambda
The Universal Lambda backend was contributed by @woodrush. Implementation details are described in the LambdaVM repository.
This backend is tested with the Universal Lambda interpreter clamb written by @irori.
Interactive programs require the -u
option which disables standard output buffering, used as clamb -u [input file]
.
The interpreter is automatically cloned and built via tools/runulamb.sh when the tests are run.
The output of this backend is an untyped lambda calculus term written in the binary lambda calculus notation. The output program is written as a sequence of 0/1s in ASCII. The bit stream must be packed into a byte stream before passing it to the interpreter. This can be done using tools/packbits.c. Please see tools/runulamb.sh for usage details.
Grass
The Grass backend was contributed by @woodrush. Implementation details are described in the GrassVM and LambdaVM repositories.
This backend is tested with the Grass interpreter grass.ml, originally written by @ytomino and modified by @youz and @woodrush. The modifications are described in the GrassVM repository.
Future works
I'm interested in
- adding more backends (e.g., 16bit CPU, Malbolge Unshackled, ...)
- running more programs (e.g., lua.bf or mruby.bf?)
- supporting more C features (e.g., bit operations)
- eliminating unnecessary code in 8cc
Adding a backend shouldn't be extremely difficult. PRs are welcomed!
See also
This project is a sequel of bflisp.
Acknowledgement
I'd like to thank Rui Ueyama for his easy-to-hack compiler and suggesting the basic idea which made this possible.